ENTOMOLOGIST AND BOTANIST. 



233 



s, the larvie and perfect insects usually 

 feed together, and it might be thought impos- 

 sible, at first sight, to tell wliich are the more 

 destructive. But when we consider the import- 

 ant fact to which we have before adverted, that 

 the whole growth of insects takes place in the 

 larva state, we must conclude that even in these 

 cases, the principal damage must be effected 

 wliilst the insects are in the prepai-atory stage. 

 If to this we add that one of the most destructive 

 orders of insects, namely, the Lepidoptera, com- 

 mit all tlieir havoc in the lai-va form of cater- 

 pillars, we shall be able to form some estimate 

 of the preponderance of damage effected by in- 

 sects in the larva state. 



Let us now inquire if lai'vas exliibit any char- 

 acters by which we can so classify them as to 

 determine to what orders and families they will 

 respectively belong when they shall have attained 

 their perfect state. 



The difficulty which has attended all attempt*; 

 to classify lai-vfe upon their own characters, and 

 at the same time preserve their relationsliip to 

 their respective imagines, strongly exlubits the 

 comparative inferiority or degi-adation of the 

 larval state. We can indeed classify lai-vse into 

 what seem to be natural groups, founded upou 

 their most important and prominent characters ; 

 but when we come to put opposite to them, in 

 parallel series, the perfect insects which these 

 lai-vas produce, we are astonished to find that 

 every vestige of relationship is lost. Take, for 

 example, the classification of lar\;c li\ Kirby and 

 Spence. These authors anani^r hirvie in five 

 principal groups. The thst uroiip produces, 

 when arrived at the perfect state, a heterogene- 

 ous mixture of Coleoptera, Hymenopt«ra and 

 Diptera. The second group produces Diptera 

 only. In the third, two of the most remote 

 orders of insects, the Coleoptera aiid Neuroptera 

 are brought into juxtaposition. In the fourth, a 

 part of the Tipulidse are separated from the rest 

 of their family, and . from the Dipterous order, 

 and associated ^vith the Micro-Lepidoptera. And 

 in the fifth group, Coleoptera, Hymenoptera and 

 Lepidoptera are indiscriminately associated to- 

 gether. 



Still the practical question remains whether 

 an}' general rules can be established, by which 

 we can know what form the noxious larvas we 

 meet with, will ultimately assume. 



In order to understand what follows, it is 

 necessary to state that larvse have legs of two 

 kinds: first, the true legs, representing the legs 

 of the perfect insect, wMch are comparatively 

 firm, conical, and jointed, and, when present, are 

 almost always six in number, and attached to 



the first three segments of the body. Secondly, 

 the spurious legs, or prolegs, which are short, 

 thick, muscular and unarticulated, varying in 

 number from two to sixteen, and attached to one 

 or more of the eight last segments. 



1st. Generalization. All larvie generally 

 known as Catcrpillai's, and ilistinguished by 

 having liotli leg- and prolegs, produce either 

 Lepiduplrra. or Saw-flies in the order Hymen- 

 optera ; and the larva; of the Saw-flies are dis- 

 tinguished from those of the Lepidoptera by 

 having more than five pairs of prolegs ; and by 

 having only two eyes, wliilst the true caterpillars 

 have ten or twelve, and also by their habit of 

 rolling themselves into a spiral coil. 



2d. As a general rule, hairy caterpillars pro- 

 duce moths, whilst spiny or naked ones produce 

 buttei-flies or spliinges. The rule may be more 

 accurately stated thus : All densely haired cater- 

 pillars produce moths, but all the larvse of moths 

 are not hairy. The caterijiUars of the butterflies 

 and spliinges are either naked or ornamented 

 with spines, or with very short or scattered 

 hairs. 



3d. Wood-boriug larvije belong mostly to the 

 Coleoptera ; but also to a few families of the 

 Lepidoptera, namely, the jEgeridas, the Hepi- 

 alid;e, and a few exceptional Tortricidas. The 

 larvje of the Lepidoptera can always be distin- 

 guished from those of the Coleoptera, by the 

 presence of prolegs on the intermediate seg- 

 ments. A few Coleopterous larvte have one pair 

 of prolegs on the anal segment, but more gener- 

 ally only one such leg. 



4th. All leaf-sucking larvfe belong to the order 

 Hemiptera (including the Homoptera). 



5th. All leaf-gnawing larviB, excepting grass- 

 hoppers, and the caterpillars above treated of, 

 belong to one tribe of Coleoptera, distinguished 

 by the title of Pliyliophaga, or Leaf-eaters, and 

 comprising the four families Criocerid», Galer- 

 ucidaj, Cassididas, and Chrysomelidis. These 

 lai-vje, moreover, can generally be identified by 

 tlieir short, wrinkled forms, their sluggish mo- 

 tions, and some of them by the singular habit of 

 protecting their bodies by their own excrement. 



6th. All lai-vae found underground, excepting 

 those which enter it only for the pm-pose of 

 undergoing their transformations, are divisible, 

 according to their habits, into two sections. 

 First, the subterranean larvae, properly so called, 

 which live under ground, and feed upon the 

 roots of plants ; and, secondly, those which sub- 

 sist above ground, but burrow into it, when not 

 feeding, for the purpose of concealment. True 

 subterranean larvse are found in the orders Cole- 

 optera, Hemiptera, Homoptera, and Diptera. 



